DanielLeens commented on PR #11206: URL: https://github.com/apache/seatunnel/pull/11206#issuecomment-5459248432
Correcting a stale CI fact from my review a few minutes ago: I said `mysql-cdc-connector-it` was "still running and hasn't produced a result yet" on this head. That run (`33139046600`) has since completed, and `mysql-cdc-connector-it` finished `failure` on both JDK 8 and JDK 11. I pulled the raw job logs rather than trusting the check name alone, because this test's failure mode has changed meaningfully from earlier rounds and it deserves a precise read rather than a generic "still red" note: - Source-side registration now works. Across every Flink minor version in the matrix (1.13/1.15/1.18/1.20), the log shows `Registered newly added CDC table source.payments_rt_<suffix>` firing right after the `ALTER TABLE` that adds the watched table -- this is real progress over the historical failure mode where that line never appeared at all. - The job still fails, but later and differently: `AbstractMysqlCDCITBase.queryNewlyAddedWildcardSinkTable` hits `java.sql.SQLSyntaxErrorException: Table 'sink.source_payments_rt_<suffix>' doesn't exist`, and the Awaitility wait for that table to become readable times out at 2 minutes. This reproduces identically on both JDK 8 and JDK 11, across all four Flink versions the job matrix runs. So this is not flaky/intermittent -- the sink-side table for a newly-discovered wildcard table is not getting created within the wait window, every time. Given that pattern, this is a genuine, still-open feature gap (the sink-side auto-create/schema-save-mode path for a newly added table isn't completing in time, or isn't being triggered at all), not CI noise -- so it does not change my Merge Recommendation. It's an addition to my existing blockers list, not a replacement: Issues 1 and 2 (High) from my review a few minutes ago -- runtime identifier validation failing open, and the missing sink-capability gate before runtime factory discovery -- remain the primary blockers on this head regardless of this CI result. (The other three failing jobs on this run -- `unit-test (11, windows-latest)` and `all-connectors-it-7` on both JDKs -- I have not dug into yet; flagging that I haven't confirmed those are unrelated so it's not misread as a clean bill of health on the rest of the run.) -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
